PROGRESS IN THE STUDY OF VARIATIOX. 61 



occurs. A considerable series of intermediates from this 

 region is preserved in the British Museum, and from these 

 and others described by Dresser it appears that here again 

 there is no definite race of intermediates grouped round 

 the mean as a normal, but that the intermediates are a 

 heterogeneous body, and highly variable. It is quite 

 possible that on seriation they would be found to be on the 

 whole divisible into two groups favouring the two types 

 respectively. 



Birds, however, are not well adapted to statistical treat- 

 ment of the kind contemplated. Their large size, fre- 

 quently migratory habits, and the comparative difficulty in 

 preserving them in large numbers makes them for the most 

 part unpromising material for such work. Besides this, it 

 is in very exceptional cases only that they can be bred in 

 captivity. Before passing to other examples, however, it 

 may be noted that in these cases, though the individual 

 parentage of the intermediates or of the types in the area 

 of intergradation is scarcely as yet a matter of observa- 

 tion, there is nothing in the facts inconsistent with the 

 possibility that the whole phenomenon may be one of dis- 

 continuous variation. Nevertheless, it is quite as possible 

 that the intermediates may in each case be the result of 

 cross-breeding. If the latter is the true view, then, since 

 the whole population is not found to have regressed to the 

 mean form, this state of things must be due either to the 

 existence of some principle by which the types mate more 

 often legitimately than illegitimately, or to the existence of 

 natural discontinuity, or to the unfitness and consequent 

 extermination of the mean forms. The last hypothesis is 

 so improbable that it need scarcely be considered. No 

 support could be produced in behalf of such a view, either 

 from facts of the case or analogy of other cases. 



Among Lepidoptera there is perhaps more hope of ob- 

 taining direct evidence on this question, and there are 

 numerous cases which would well repay careful work. 

 Scarcely any have been touched as yet. 



A remarkable example has been observed by W. H. 

 Edwards (4) among the N. American Papilios. The form 



